bolt.diy/app/utils/fileExtract.ts

import * as pdfjsLib from 'pdfjs-dist/build/pdf.mjs';
import pdfWorker from 'pdfjs-dist/build/pdf.worker.mjs';
import mammoth from 'mammoth/mammoth.browser';

pdfjsLib.GlobalWorkerOptions.workerSrc = pdfWorker;

export async function extractTextFromFile(file: File): Promise<string> {
  if (file.type === 'application/pdf' || file.name.toLowerCase().endsWith('.pdf')) {
    const arrayBuffer = await file.arrayBuffer();
    const pdf = await pdfjsLib.getDocument({ data: arrayBuffer }).promise;
    const texts: string[] = [];

    for (let i = 1; i <= pdf.numPages; i++) {
      const page = await pdf.getPage(i);
      const content = await page.getTextContent();
      texts.push(content.items.map((item: any) => item.str).join(' '));
    }

    return texts.join('\n');
  }

  if (
    file.type === 'application/vnd.openxmlformats-officedocument.wordprocessingml.document' ||
    file.name.toLowerCase().endsWith('.docx')
  ) {
    const arrayBuffer = await file.arrayBuffer();
    const result = await mammoth.extractRawText({ arrayBuffer });

    return result.value;
  }

  return file.text();
}